U.S. patent number 10,142,140 [Application Number 15/421,295] was granted by the patent office on 2018-11-27 for apparatus for receiving signal based on faster-than-nyquist and method for using the same.
This patent grant is currently assigned to ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE. The grantee listed for this patent is ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE. Invention is credited to Myung-Sun Baek, Bi-Woong Chung, Nam-Ho Hur, Nam-Shik Kim, Sang-Woon Kwak, Hae-Chan Kwon, Hyoung-Soo Lim, Joung-Il Yun.
United States Patent |
10,142,140 |
Yun , et al. |
November 27, 2018 |
Apparatus for receiving signal based on faster-than-Nyquist and
method for using the same
Abstract
Disclosed herein are an apparatus and method for receiving a
signal based on FTN. The apparatus for receiving a signal based on
FTN includes an equalizer for creating a Log Likelihood Ratio (LLR)
sequence by equalizing an FTN signal sequence sampled at an FTN
signaling rate; a deinterleaver for deinterleaving the created LLR
sequence; a decoder for decoding the LLR sequence by correcting
errors in the deinterleaved LLR sequence; an interleaver for
interleaving the decoded LLR sequence and providing the interleaved
LLR sequence to the equalizer; and an FTN interference estimation
unit for providing the FTN signal sequence, from which an FTN
interference sequence is eliminated, to the equalizer, using the
interleaved LLR sequence.
Inventors: |
Yun; Joung-Il (Daejeon,
KR), Kwak; Sang-Woon (Daejeon, KR), Baek;
Myung-Sun (Daejeon, KR), Kwon; Hae-Chan (Daejeon,
KR), Lim; Hyoung-Soo (Daejeon, KR), Hur;
Nam-Ho (Sejong, KR), Kim; Nam-Shik (Hwaseong,
KR), Chung; Bi-Woong (Yongin, KR) |
Applicant: |
Name |
City |
State |
Country |
Type |
ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE |
Daejeon |
N/A |
KR |
|
|
Assignee: |
ELECTRONICS AND TELECOMMUNICATIONS
RESEARCH INSTITUTE (Daejeon, KR)
|
Family
ID: |
61010263 |
Appl.
No.: |
15/421,295 |
Filed: |
January 31, 2017 |
Prior Publication Data
|
|
|
|
Document
Identifier |
Publication Date |
|
US 20180034591 A1 |
Feb 1, 2018 |
|
Foreign Application Priority Data
|
|
|
|
|
Jul 29, 2016 [KR] |
|
|
10-2016-0097106 |
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
H04L
27/01 (20130101); H04L 1/0071 (20130101); H04L
1/0057 (20130101); H04L 1/0047 (20130101); H04L
25/068 (20130101); H04L 25/03834 (20130101); H04L
25/03006 (20130101); H04L 1/005 (20130101); H04L
1/0054 (20130101); H04B 1/71072 (20130101) |
Current International
Class: |
H04L
1/00 (20060101); H04L 25/06 (20060101); H04L
27/01 (20060101); H04B 1/7107 (20110101); H04L
25/03 (20060101) |
References Cited
[Referenced By]
U.S. Patent Documents
Foreign Patent Documents
|
|
|
|
|
|
|
10-2015-0097048 |
|
Aug 2015 |
|
KR |
|
Other References
Baek et al. "An Efficient Turbo Equalization for Faster than
Nyquist Signal", pp. 231-234, International Journal of Signal
Processing Systems vol. 4, No. 3, Jun. 2016. cited by examiner
.
Anderson et al. "Turbo Equalization and an M-BCJR Algorithm for
Strongly Narrowband Intersymbol Interference", pp. 262-266,
ISITA2010, Taichung, Taiwan, Oct. 17-20, 2010. cited by examiner
.
Maalouli et al., "Performance Analysis of a MMSE Turbo Equalizer
with LDPC in a FTN Channel with Application to Digital Video
Broadcast", pp. 1871-1875, Asilomar 2014. cited by examiner .
Prlja et al. "Receivers for Faster-than-Nyquist Signaling with and
without Turbo Equalization", pp. 464-468, ISIT 2008, Toronto,
Canada, Jul. 6-11, 2008. cited by examiner .
Sen et al. "A Low-Complexity Graph-Based LMMSE Receiver Designed
for Colored Noise Induced by FTN-Signaling", pp. 642-647 IEEE 2014.
cited by examiner.
|
Primary Examiner: Vlahos; Sophia
Claims
What is claimed is:
1. An apparatus for receiving a signal based on Faster-Than-Nyquist
(FTN), comprising: an equalizer for creating a Log Likelihood Ratio
(LLR) sequence by equalizing an FTN signal sequence sampled at an
FTN signaling rate; a deinterleaver for deinterleaving the created
LLR sequence; a decoder for decoding the LLR sequence by correcting
errors in the deinterleaved LLR sequence; an interleaver for
interleaving the decoded LLR sequence and providing the interleaved
LLR sequence to the equalizer; and an FTN interference estimation
unit for providing the FTN signal sequence, from which an FTN
interference sequence is eliminated, to the equalizer, using the
interleaved LLR sequence.
2. The apparatus of claim 1, wherein the FTN interference
estimation unit comprises: a modulation unit for estimating a
symbol sequence of the FTN signal sequence by modulating the
interleaved LLR sequence; and an FTN interference filter for
estimating the FTN interference sequence using the estimated symbol
sequence, for eliminating the estimated FTN interference sequence
from the FTN signal sequence, and for providing the equalizer with
the FTN signal sequence from which the estimated FTN interference
sequence is eliminated.
3. The apparatus of claim 2, wherein the FTN interference filter
uses FTN interference tap coefficients, and wherein the FTN
interference filter sets the FTN interference tap coefficients,
which are incorporated in the equalizer in order to reconstruct
symbols, to `0`.
4. The apparatus of claim 3, wherein the FTN interference filter
estimates the FTN interference sequence by performing convolution
of the estimated symbol sequence with the FTN interference tap
coefficients.
5. The apparatus of claim 4, wherein the FTN interference filter
eliminates the estimated FTN interference sequence from the FTN
signal sequence, thereby eliminating an FTN interference component,
due to the FTN interference tap coefficients that are not
incorporated in the equalizer, from the FTN signal sequence.
6. The apparatus of claim 5, wherein the FTN interference filter
iterates a demodulation and decoding process until a result of
subtracting the estimated symbol sequence from the symbol sequence
of the FTN signal sequence satisfies a preset condition, thereby
eliminating the FTN interference component, due to the FTN
interference tap coefficients that are not incorporated in the
equalizer, from the FTN signal sequence.
7. The apparatus of claim 6, wherein the decoder is configured to:
output the decoded LLR sequence as an information bit sequence when
the result of eliminating the estimated FTN interference sequence
from the FTN signal sequence satisfies a preset condition, and
provide the decoded LLR sequence to the interleaver so as to
iterate the demodulation and decoding process when the result of
eliminating the estimated FTN interference sequence from the FTN
signal sequence does not satisfy the preset condition.
8. The apparatus of claim 7, wherein the equalizer determines a
range within which the FTN interference tap coefficients fall in
consideration of a number of iterations of the demodulation and
decoding process.
9. The apparatus of claim 8, wherein the equalizer is a BCJR
equalizer for implementing a Bahl-Cocke-Jelinek-Raviv (BCJR)
algorithm and wherein the decoder is an LDPC decoder for performing
Low-Density Parity-Check (LDPC) decoding, the deinterleaver
deinterleaves an extrinsic information sequence of the equalizer,
acquired by eliminating an extrinsic information sequence of the
decoder, calculated by the interleaver, from the LLR sequence
created by the BCJR equalizer.
10. The apparatus of claim 9, wherein the equalizer is the BCJR
equalizer and wherein the decoder is the LDPC decoder, the
interleaver provides the BCJR equalizer with the extrinsic
information sequence of the decoder, acquired by eliminating the
extrinsic information sequence of the equalizer from the LLR
sequence, which is interleaved after being LDPC-decoded by the LDPC
decoder.
11. A method for receiving a signal based on Faster-Than-Nyquist
(FTN), using an apparatus for receiving a signal based on FTN,
comprising: creating, by an equalizer, a Log Likelihood Ratio (LLR)
sequence by equalizing an FTN signal sequence sampled at an FTN
signaling rate; deinterleaving, by a deinterleaver, the created LLR
sequence; decoding, by a decoder, the LLR sequence by correcting
errors in the deinterleaved LLR sequence; interleaving, by an
interleaver, the decoded LLR sequence; and providing, by an FTN
interference estimation unit, the FTN signal sequence from which an
FTN interference sequence is eliminated to the equalizer, using the
interleaved LLR sequence, wherein the FTN interference estimation
units uses FTN interference tap coefficients to provide the FTN
interference sequence, and the FTN interference tap coefficients
corresponding to tap coefficients used by the equalizer to create
the LLR sequence are set to `0`.
12. The method of claim 11, wherein providing the FTN signal
sequence comprises: estimating a symbol sequence of the FTN signal
sequence by modulating the interleaved LLR sequence; estimating the
FTN interference sequence using the estimated symbol sequence;
eliminating the estimated FTN interference sequence from the FTN
signal sequence; and providing the equalizer with the FTN signal
sequence from which the estimated FTN interference sequence is
eliminated.
13. A method for receiving a signal based on Faster-Than-Nyquist
(FTN), using an apparatus for receiving a signal based on FTN, the
method comprising: creating, by an equalizer, a Log Likelihood
Ratio (LLR) sequence by equalizing an FTN signal sequence sampled
at an FTN signaling rate; deinterleaving, by a deinterleaver, the
created LLR sequence; decoding, by a decoder, the LLR sequence by
correcting errors in the deinterleaved LLR sequence; interleaving,
by an interleaver, the decoded LLR sequence; and providing, by an
FTN interference estimation unit, the FTN signal sequence from
which an FTN interference sequence is eliminated to the equalizer,
using the interleaved LLR sequence, wherein providing the FTN
signal sequence comprises: estimating a symbol sequence of the FTN
signal sequence by modulating the interleaved LLR sequence;
estimating the FTN interference sequence using the estimated symbol
sequence; eliminating the estimated FTN interference sequence from
the FTN signal sequence, and providing the equalizer with the FTN
signal sequence from which the estimated FTN interference sequence
is eliminated, and wherein estimating the FTN interference sequence
comprises setting FTN interference tap coefficients used to
estimate the FTN interference sequence, which are incorporated in
the equalizer in order to reconstruct symbols, to `0`.
14. The method of claim 13, wherein estimating the FTN interference
sequence comprises estimating the FTN interference sequence by
performing convolution of the estimated symbol sequence with the
FTN interference tap coefficients.
15. The method of claim 14, wherein eliminating the estimated FTN
interference sequence comprises eliminating the estimated FTN
interference sequence from the FTN signal sequence, whereby an FTN
interference component, caused due to the FTN interference tap
coefficients that are not incorporated in the equalizer, is
eliminated.
16. The method of claim 15, wherein eliminating the estimated FTN
interference sequence comprises iterating a demodulation and
decoding process until a result of subtracting the estimated symbol
sequence from the symbol sequence of the FTN signal sequence
satisfies a preset condition, whereby the FTN interference
component, caused due to the FTN interference tap coefficients that
are not incorporated in the equalizer, is eliminated from the FTN
signal sequence.
17. The method of claim 16, wherein decoding the LLR sequence
comprises outputting the decoded LLR sequence as an information bit
sequence when a result of eliminating the estimated FTN
interference sequence from the FTN signal sequence satisfies a
preset condition.
18. The method of claim 17, wherein decoding the LLR sequence
comprises providing the decoded LLR sequence to the interleaver so
as to iterate the demodulation and decoding process when the result
of eliminating the estimated FTN interference sequence from the FTN
signal sequence does not satisfy the preset condition.
19. The method of claim 18, wherein the equalizer is a BCJR
equalizer for implementing a Bahl-Cocke-Jelinek-Raviv (BCJR)
algorithm and wherein the decoder is an LDPC decoder for performing
Low-Density Parity-Check (LDPC) decoding, deinterleaving the
created LLR sequence comprises deinterleaving an extrinsic
information sequence of the equalizer, acquired by eliminating an
extrinsic information sequence of the decoder, calculated by the
interleaver, from the LLR sequence created by the BCJR
equalizer.
20. The method of claim 19, wherein the equalizer is the BCJR
equalizer and wherein the decoder is the LDPC decoder, interleaving
the decoded LLR sequence comprises providing the BCJR equalizer
with the extrinsic information sequence of the decoder, acquired by
eliminating the extrinsic information sequence of the equalizer
from the interleaved LLR sequence.
Description
CROSS REFERENCE TO RELATED APPLICATION
This application claims the benefit of Korean Patent Application
No. 10-2016-0097106, filed Jul. 29, 2016, which is hereby
incorporated by reference in its entirety into this
application.
BACKGROUND OF THE INVENTION
1. Technical Field
The present invention relates generally to broadcasting and
communications technology, and more particularly to technology for
receiving signals based on Faster-Than-Nyquist (FTN).
2. Description of the Related Art
General communications systems use a Nyquist pulse-shaping method
in order to transmit signals without interference between symbols.
The Nyquist pulse-shaping method is a pulse-shaping method that is
capable of achieving the maximum transmission rate in a given
bandwidth without interference between symbols. However, in recent
communication systems, requirements for higher spectral efficiency
are increasing, but the Nyquist pulse-shaping method has a
limitation as to transmission efficiency. Accordingly, as a method
for improving this, the Faster-Than-Nyquist (FTN) method has been
introduced.
In the FTN method, a pulse shape, given depending on frequency
bandwidth, is kept, but the time interval between shaping of a
pulse and shaping of the next pulse is decreased. That is, the gap
between symbols is narrowed, whereby a signal, the pulses of which
overlap each other, is transmitted. Accordingly, the FTN method
necessarily causes Inter-Symbol Interference (ISI), but may have a
higher signal transfer rate than the Nyquist pulse-shaping method
for the same bandwidth.
As described above, the FTN method may improve transmission speed,
but ISI, which is not caused in the Nyquist method, is included in
a signal during the process of generating the signal and is
transmitted when the signal is transmitted. Therefore, a receiver
needs to eliminate ISI (hereinafter, referred to as "FTN
interference") caused by FTN in order to reconstruct data without
errors.
Because FTN interference is intentionally generated in the process
of generating a signal to be transmitted, the pattern is accurately
known, and thus FTN interference may be eliminated from the
received signal. That is, a receiver generates a reference signal
in which original data includes intentional interference caused by
FTN, the received signal is compared with the reference signal, and
thereby data from which the interference is eliminated is
reconstructed. Also, if a pulse-shaping filter has a large enough
number of taps to generate a bandlimited signal, FTN interference
generated through such long filter taps causes interference between
neighboring symbols in a wider range. Here, if a receiver cannot
sufficiently eliminate the interference, it is difficult to improve
reception performance, but if the range of the interference between
neighboring symbols to be processed is wider, reception performance
is improved, but complexity is increased.
Meanwhile, Korean Patent Application Publication No.
10-2015-0097048, titled "Signal-receiving apparatus based on
Faster-Than-Nyquist and signal-decoding method thereof relates to
an apparatus for receiving FTN-based signals and a method for
decoding FTN-based signals. This patent discloses a
signal-receiving apparatus based on FTN, which includes an
equalizer for calculating, when a signal sampled with
Faster-Than-Nyquist (FTN) is received on a communication channel, a
posterior probability of information bits for the received signal
through the BCJR algorithm and for calculating a Log Likelihood
Ratio (LLR) using the calculated posterior probability; a
deinterleaver for deinterleaving bit data output from the
equalizer; a decoder for compensating for signal interference of
the data bits deinterleaved by the deinterleaver using the LLR,
thereby decoding the data; and an interleaver for interleaving the
data output from the decoder and providing the interleaved data to
the equalizer.
However, Korean Patent Application Publication No. 10-2015-0097048
does not mention a problem related to FTN interference occurring at
a receiver of FTN-based signals.
SUMMARY OF THE INVENTION
An object of the present invention is to improve the equalization
performance of an equalizer having relatively low complexity in a
digital communication system using FTN.
Another object of the present invention is to accurately detect
transmitted signals in an equalization process having relatively
low complexity in a digital communication system using FTN.
A further object of the present invention is to effectively receive
FTN signals by eliminating interference components that cannot be
cancelled in an equalizer due to the complexity problem.
Yet another object of the present invention is to reduce the amount
of resources for an interleaver when a signal receiver using a
Bahl-Cocke-Jelinek-Raviv (BCJR) equalizer and a Low-Density
Parity-Check (LDPC) decoder is implemented.
Still another object of the present invention is to improve the
equalization performance of a BCJR equalizer by eliminating FTN
interference components caused due to interference tap coefficients
that are not incorporated in the BCJR equalizer, and to accurately
receive FTN signals by improving the error-correction capability of
an LDPC decoder.
In order to accomplish the above objects, an apparatus for
receiving a signal based on Faster-Than-Nyquist (FTN) according to
an embodiment of the present invention includes an equalizer for
creating a Log Likelihood Ratio (LLR) sequence by equalizing an FTN
signal sequence sampled at an FTN signaling rate; a deinterleaver
for deinterleaving the created LLR sequence; a decoder for decoding
the LLR sequence by correcting errors in the deinterleaved LLR
sequence; an interleaver for interleaving the decoded LLR sequence
and providing the interleaved LLR sequence to the equalizer; and an
FTN interference estimation unit for providing the FTN signal
sequence, from which an FTN interference sequence is eliminated, to
the equalizer, using the interleaved LLR sequence.
Here, the FTN interference estimation unit may include a modulation
unit for estimating a symbol sequence of the FTN signal sequence by
modulating the interleaved LLR sequence; and an FTN interference
filter for estimating the FTN interference sequence using the
estimated symbol sequence, for eliminating the estimated FTN
interference sequence from the FTN signal sequence, and for
providing the equalizer with the FTN signal sequence from which the
estimated FTN interference sequence is eliminated.
Here, the FTN interference filter may set FTN interference tap
coefficients, which are incorporated in the equalizer in order to
reconstruct symbols, to `0` in an FTN interference filter tap
coefficient sequence of the FTN signal sequence, using the
estimated symbol sequence.
Here, the FTN interference filter may estimate the FTN interference
sequence by performing convolution of the estimated symbol sequence
with the FTN interference filter tap coefficient sequence.
Here, the FTN interference filter may eliminate the estimated FTN
interference sequence from the FTN signal sequence, thereby
eliminating an FTN interference component, caused due to the FTN
interference tap coefficients that are not incorporated in the
equalizer, from the FTN signal sequence.
Here, the FTN interference filter may iterate a demodulation and
decoding process until a result of subtracting the estimated symbol
sequence from the symbol sequence of the FTN signal sequence
satisfies a preset condition, thereby eliminating the FTN
interference component, caused due to the FTN interference tap
coefficients that are not incorporated in the equalizer, from the
FTN signal sequence.
Here, the decoder may output the decoded LLR sequence as an
information bit sequence when the result of eliminating the
estimated FTN interference sequence from the FTN signal sequence
satisfies a preset condition.
Here, the decoder may provide the decoded LLR sequence to the
interleaver so as to iterate the demodulation and decoding process
when the result of eliminating the estimated FTN interference
sequence from the FTN signal sequence does not satisfy the preset
condition.
Here, the equalizer may determine a range within which the FTN
interference tap coefficients fall in consideration of a number of
iterations of the demodulation and decoding process.
Here, when the equalizer is a BCJR equalizer for implementing a
Bahl-Cocke-Jelinek-Raviv (BCJR) algorithm and when the decoder is
an LDPC decoder for performing Low-Density Parity-Check (LDPC)
decoding, the deinterleaver may deinterleave an extrinsic
information sequence of the equalizer, acquired by eliminating an
extrinsic information sequence of the decoder, calculated by the
interleaver, from the LLR sequence created by the BCJR
equalizer.
Here, when the equalizer is the BCJR equalizer and when the decoder
is the LDPC decoder, the interleaver may provide the BCJR equalizer
with the extrinsic information sequence of the decoder, acquired by
eliminating the extrinsic information sequence of the equalizer
from the interleaved LLR sequence.
Also, in order to accomplish the above objects, a method for
receiving a signal based on Faster-Than-Nyquist (FTN), using an
apparatus for receiving a signal based on FTN, according to an
embodiment of the present invention, includes creating, by an
equalizer, a Log Likelihood Ratio (LLR) sequence by equalizing an
FTN signal sequence sampled at an FTN signaling rate;
deinterleaving, by a deinterleaver, the created LLR sequence;
decoding, by a decoder, the LLR sequence by correcting errors in
the deinterleaved LLR sequence; interleaving, by an interleaver,
the decoded LLR sequence; and providing, by an FTN interference
estimation unit, the FTN signal sequence from which a part of an
FTN interference sequence is eliminated to the equalizer, using the
interleaved LLR sequence.
Here, providing the FTN signal sequence may include estimating a
symbol sequence of the FTN signal sequence by modulating the
interleaved LLR sequence; estimating the FTN interference sequence
using the estimated symbol sequence; eliminating the estimated FTN
interference sequence from the FTN signal sequence; and providing
the equalizer with the FTN signal sequence from which the estimated
FTN interference sequence is eliminated.
Here, estimating the FTN interference sequence may be configured to
set FTN interference tap coefficients, which are incorporated in
the equalizer in order to reconstruct symbols, to `0` in an FTN
interference filter tap coefficient sequence of the FTN signal
sequence, using the estimated symbol sequence.
Here, estimating the FTN interference sequence may be configured to
estimate the FTN interference sequence by performing convolution of
the estimated symbol sequence with the FTN interference filter tap
coefficient sequence.
Here, eliminating the estimated FTN interference sequence may be
configured to eliminate the estimated FTN interference sequence
from the FTN signal sequence, whereby an FTN interference
component, caused due to the FTN interference tap coefficients that
are not incorporated in the equalizer, may be eliminated.
Here, eliminating the estimated FTN interference sequence may be
configured to iterate a demodulation and decoding process until a
result of subtracting the estimated symbol sequence from the symbol
sequence of the FTN signal sequence satisfies a preset condition,
whereby the FTN interference component that is not incorporated in
the equalizer may be eliminated from the FTN signal sequence.
Here, decoding the LLR sequence may be configured to output the
decoded LLR sequence as an information bit sequence when a result
of eliminating the estimated FTN interference sequence from the FTN
signal sequence satisfies a preset condition.
Here, decoding the LLR sequence may be configured to provide the
decoded LLR sequence to the interleaver so as to iterate the
demodulation and decoding process when the result of eliminating
the estimated FTN interference sequence from the FTN signal
sequence does not satisfy the preset condition.
Here, when the equalizer is a BCJR equalizer for implementing a
Bahl-Cocke-Jelinek-Raviv (BCJR) algorithm and when the decoder is
an LDPC decoder for performing Low-Density Parity-Check (LDPC)
decoding, deinterleaving the created LLR sequence may be configured
to deinterleave an extrinsic information sequence of the equalizer,
acquired by eliminating an extrinsic information sequence of the
decoder, calculated by the interleaver, from the LLR sequence
created by the BCJR equalizer.
Here, when the equalizer is the BCJR equalizer and when the decoder
is the LDPC decoder, interleaving the decoded LLR sequence may be
configured to provide the BCJR equalizer with the extrinsic
information sequence of the decoder, acquired by eliminating the
extrinsic information sequence of the equalizer from the
interleaved LLR sequence.
BRIEF DESCRIPTION OF THE DRAWINGS
The above and other objects, features and advantages of the present
invention will be more clearly understood from the following
detailed description taken in conjunction with the accompanying
drawings, in which:
FIG. 1 is a view that shows an FTN-based signal reception model
according to an embodiment of the present invention;
FIGS. 2A to 2C are graphs that show FTN interference filter tap
coefficients depending on an FTN factor .tau. according to an
embodiment of the present invention;
FIG. 3 is a block diagram of an apparatus for receiving an
FTN-based signal according to an embodiment of the present
invention;
FIG. 4 is a block diagram that specifically shows an example of the
FTN interference estimation unit illustrated in FIG. 3;
FIGS. 5 and 6 are block diagrams that show an FTN-based signal
reception apparatus using a BCJR equalizer and an LDPC decoder
according to an embodiment of the present invention;
FIG. 7 is a flowchart that shows a method for receiving an
FTN-based signal according to an embodiment of the present
invention; and
FIG. 8 is a flowchart that specifically shows an example of the
step of eliminating an FTN interference sequence, illustrated in
FIG. 7.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
The present invention will be described in detail below with
reference to the accompanying drawings. Repeated descriptions and
descriptions of known functions and configurations which have been
deemed to make the gist of the present invention unnecessarily
obscure will be omitted below. The embodiments of the present
invention are intended to fully describe the present invention to a
person having ordinary knowledge in the art to which the present
invention pertains. Accordingly, the shapes, sizes, etc. of
components in the drawings may be exaggerated in order to make the
description clearer.
Hereinafter, a preferred embodiment according to the present
invention will be described in detail with reference to the
accompanying drawings.
FIG. 1 is a view that shows an FTN-based signal reception model
according to an embodiment of the present invention.
Referring to FIG. 1, in the FTN-based signal reception model
according to an embodiment of the present invention, u={u.sub.0,
u.sub.1, . . . , u.sub.(K-1)}, which is the information bit
sequence to be transmitted and which has a length of K, is
channel-coded with a coding rate R, whereby c={c.sub.0, c.sub.1. .
. , c.sub.(N-1)}, which is a bit sequence having a length of N=K/R,
may be acquired. The bit sequence c may be changed to v={v.sub.0,
v.sub.1, . . . , v.sub.(N-1)} by changing the order in which the
elements in c are arranged through an interleaving process. Then,
the bit sequence v is mapped to a={a.sub.0, a.sub.1, . . . ,
a.sub.(L-1)}, which is a symbol sequence having a length of L=N/m,
through a modulation process (here, M-ary modulation may be used
according to an embodiment of the present invention, thus
m=log.sub.2M may be used), and a signal p(t), in which each symbol
in the symbol sequence is output every .tau.T seconds, may be
represented as the following Equation (1):
.function..times..times..delta..function..times..times..tau..times..times-
. ##EQU00001##
where a.sub.1 denotes the l-th symbol in the symbol sequence a, and
.delta.(t) denotes the Dirac delta function.
Also, T is a symbol period when a transmission filter for limiting
the band of p(t) satisfies the Nyquist criterion, and .tau. denotes
an FTN factor. Here, the FTN factor .tau. in the range of
0<.tau..ltoreq.1 may be used, and the transmission speed may
increase in proportion to 1/.tau.. Particularly, when .tau.=1, the
transmission period may satisfy the Nyquist rate.
The modulated signal p(t) may pass through a transmission filter
having an impulse response h(t) and may then be transmitted, and
the transmitted signal x(t) may be represented as the following
Equation (2):
.function..function..function..times..times..function..times..times..tau.-
.times..times. ##EQU00002##
Here, the transmission filter h(t) uses a Root Raised Cosine (RRC)
filter. Also, according to an embodiment of the present invention,
an * operator may be a convolution operator.
When the signal x(t) is transmitted through an Additive White
Gaussian Noise (AWGN) channel, the received signal r(t) may pass
through a filter matched with the transmitted signal in a filtering
process at a receiver and be output as y(t), as shown in the
following Equation (3):
.function..function..omega..function..function..function..function..omega-
..function..times..times..function..times..times..tau..times..times..omega-
..function. ##EQU00003##
Here, because g(t) corresponds to h(t)*h.sup.*(-t), and because
h(t) is the impulse response of the RRC filter, g(t) may become the
impulse response of a Raised Cosine (RC) filter. Also, .omega.(t)
may correspond to the Additive White Gaussian Noise, and {tilde
over (.omega.)}(t), which is a noise signal filtered by a reception
filter, may have a colored noise feature in the case of FTN.
When a signal y(k.tau.T), acquired by synchronizing with a symbol
position and sampling y(t) with the .tau.T period, is defined as
y.sub.k, y.sub.k may be represented as the following Equation
(4):
.times..times..function..times..tau..times..times..omega..function..times-
..times..tau..times..times..times..times..omega. ##EQU00004##
where a.sub.k is a symbol that is desired to be received, and
{tilde over (.omega.)}.sub.k may be the noise included in the
sampled signal y.sub.k at t=k.tau.T.
Also, g(k.tau.T) is defined as g.sub.k, and is called an FTN
interference tap coefficient according to an embodiment of the
present invention. When Equation (4) is rearranged from the view of
reception of the symbol a.sub.k, it may be defined as the following
Equation (5):
.times..omega..times..times..times..times..noteq..times..times..ltoreq..l-
toreq. ##EQU00005##
In the above equation, z.sub.k is the FTN interference included in
the signal y.sub.k sampled at t-k.tau.T, and z.sub.k may cause an
error when receiving the symbol a.sub.k. That is, when .tau.=1,
g.sub.k is a value for sampling g(t) with a Nyquist transmission
period T, and because all the FTN interference tap coefficients,
exclusive of g.sub.0 (1 when g(t) is an RC filter), are 0, there is
no FTN interference z.sub.k.
However, when 0<.tau.<1, because most of the FTN interference
filter tap coefficients g.sub.k have values other than 0, the FTN
interference z.sub.k is added to the signal y.sub.k, thus causing
reception error.
Therefore, when a signal based on FTN is received, it is necessary
to eliminate such an FTN interference component, whereby the
transmitted symbols may be correctly detected.
FIGS. 2A to 2C are graphs that show FTN interference filter tap
coefficients based on an FTN factor .tau. according to an
embodiment of the present invention.
FIG. 2A shows the FTN interference filter tap coefficients g.sub.k
when .tau.=1. It is confirmed that all FTN interference filter tap
coefficients are 0, exclusive of the case in which g.sub.k is
g.sub.0.
FIG. 2B shows the FTN interference filter tap coefficients g.sub.k
when .tau.=0.9. It is confirmed that FTN interference filter tap
coefficients have a value other than 0 in the range from g.sub.-g
to g.sub.g, and this may cause a reception error.
FIG. 2C shows the FTN interference filter tap coefficients g.sub.k
when .tau.=0.8. It shows that the value of g.sub.k is greatly
increased compared to the case in FIG. 2B, and thus it is confirmed
that the incidence of reception error is increased as .tau. is
closer to 0.
FIG. 3 is a block diagram that shows an apparatus for receiving a
signal based on FTN according to an embodiment of the present
invention. FIG. 4 is a block diagram that specifically shows an
example of the FTN interference estimation unit illustrated in FIG.
3.
Referring to FIG. 3, an apparatus for receiving a signal based on
FTN according to an embodiment of the present invention includes an
equalizer 110, a deinterleaver 120, a decoder 130, an interleaver
140, and an FTN interference estimation unit 150.
The equalizer 110 may reconstruct distorted symbols by equalizing
an FTN signal sequence y, which is sampled at the
Faster-Than-Nyquist (FTN) signaling rate by being output from a
reception filter, using a given FTN interference pattern, and may
create L({tilde over (v)}), which is a Log Likelihood Ratio (LLR)
sequence of soft information about the received bits, by
demodulating the reconstructed symbols.
Here, L({tilde over (v)}) may be the LLR sequence of the received
bits corresponding to .nu., which is the bit sequence output from
the interleaving process illustrated in FIG. 1.
The deinterleaver 120 performs a deinterleaving process for
restoring the order in which the elements of the LLR sequence
L({tilde over (v)}), created by the equalizer 110, are arranged,
through the reverse process of interleaving at the transmitter, and
may thereby output the deinterleaved LLR sequence L({tilde over
(c)}).
The decoder 130 may output a decoded LLR sequence L( ) by
correcting errors in the deinterleaved LLR sequence L({tilde over
(c)}) through a channel decoding process.
The interleaver 140 performs an interleaving process on the decoded
LLR sequence L( ) in the same order as the interleaving process at
the transmitter, and may thereby output an interleaved LLR sequence
L( ).
For iterative demodulation and decoding, the equalizer 110 applies
the LLR sequence L({hacek over (v)}) interleaved with a priori
probability in the equalization and demodulation process, whereby
more accurate soft information may be acquired. The decoder 130 may
acquire more accurate probability information about the received
bits by repeatedly performing the demodulation and decoding
process, and may output an information bit sequence , the error of
which is corrected as much as possible, when a preset condition
related to iteration is satisfied.
However, in the process of reconstructing symbols distorted as a
result of the FTN interference, there is a problem in which a
sufficient number of FTN interference filter tap coefficients is
not incorporated due to the complexity of the equalization process.
For example, when the equalizer 110 reconstructs a.sub.k from
y.sub.k by incorporating only some FTN interference filter tap
coefficients , {g.sub.-J, . . . , 0, . . . , g.sub.J}, noise and
the FTN interference components caused by the tap coefficients { .
. . , g.sub.-(J+2), g.sub.-(J+1), g.sub.J+1), g.sub.(J+2), . . . },
which have not been incorporated in the equalizer 110, are not
eliminated. Accordingly, the correct information bit sequence may
not be received even if the signal is repeatedly equalized,
demodulated and decoded. This may be arranged by changing the above
Equation (5) to the following Equation (6):
.times..noteq..ltoreq..ltoreq.
.times..times..times..times..times..times..times..times.<>
.times..times..times..times..times..times..times..times..times..times..om-
ega. ##EQU00006##
In order to solve the above problem, the FTN interference
estimation unit 150 may provide the equalizer 110 with an FTN
signal sequence from which the FTN interference sequence is
eliminated using the interleaved LLR sequence L({hacek over
(v)}).
Referring to FIG. 4, the FTN interference estimation unit 150 may
include a modulation unit 151 and an FTN interference filter
152.
The modulation unit 151 may estimate a symbol sequence {hacek over
(a)} by modulating the interleaved LLR sequence L({hacek over (v)})
in the same manner as the modulation process at the
transmitter.
Here, the modulation unit 151 may use an M-ary modulation method,
and may estimate the length of the symbol sequence using
m=log.sub.2M.
The FTN interference filter 152 calculates the FTN interference
components caused due to the FTN interference filter tap
coefficients that are not incorporated in the equalizer 110, and
thereby acquires the estimated FTN interference sequence as the
following Equation (7): ={hacek over (a)}*{hacek over (g)} (7)
Here, {hacek over (g)} indicates that the tap coefficients
incorporated in the equalizer 110 are set to `0`, among the FTN
interference filter tap coefficients. The FTN interference filter
152 may set the FTN interference tap coefficients, incorporated in
the equalizer in order to reconstruct symbols, to `0` using the
estimated symbol sequence {hacek over (a)}.
Here, as shown in Equation (7), the FTN interference filter 152 may
estimate the FTN interference sequence by performing convolution of
the estimated symbol sequence {hacek over (a)} with {hacek over
(g)}, in which the FTN interference tap coefficients, incorporated
in the equalizer in order to reconstruct symbols, are set to
`0`.
When the equalizer 110 incorporates only the tap coefficients
{g.sub.-J, . . . , 0, . . . , g.sub.J} for the reconstruction of
symbols, {hacek over (g)}.sub.k, which is the k-th FTN interference
filter tap coefficient of {hacek over (g)}, may be represented as
the following Equation (8):
.ltoreq..ltoreq. ##EQU00007##
Also, the FTN interference filter 152 eliminates the estimated FTN
interference sequence from the received FTN signal sequence y,
thereby creating an FTN signal sequence from which the estimated
FTN interference sequence is eliminated, that is, y- , using the
summing node 155. Also, the FTN interference filter 152 using the
summing node 155 may provide y- to the equalizer 110.
Here, the equation for the k-th element of the FTN signal sequence
from which the estimated interference sequence is eliminated may be
represented as the following Equation (9):
.ltoreq..ltoreq. <>.times..times.<>
<>.times..times..omega. ##EQU00008##
The expression <1> in Equation (9) is a part that may be
reconstructed using the FTN interference tap coefficients
incorporated in the equalizer 110, and the expression <2> is
a part in which the FTN interference component caused by the FTN
interference tap coefficients that are not incorporated in the
equalizer 110 is eliminated using the estimated symbol sequence.
The expression <2> may approach 0 as a.sub.l becomes equal to
{hacek over (a)}.sub.l. Here, {hacek over (a)}.sub.l is an
estimated symbol, modulated by the modulation unit 151, and if the
result of channel decoding is accurate, {hacek over (a)}.sub.l
becomes equal to a.sub.l. Accordingly, when the result of channel
decoding becomes accurate through the iterative demodulation and
decoding process, the expression <2> may converge on 0.
In other words, through the process of eliminating the estimated
FTN interference sequence , acquired through FTN interference
estimation in the iterative demodulation and decoding structure,
from the FTN signal sequence y sampled at the FTN signaling rate,
the FTN interference estimation unit 150 may eliminate interference
that cannot be incorporated in the equalizer 110 due to a
complexity problem.
Here, the equalizer 110 may use a variable interference tap, which
changes J, which is used to determine the range of the FTN
interference tap coefficients incorporated in the equalizer 110,
depending on the number of iterations of demodulation and
decoding.
That is, the equalizer 110 may eliminate the FTN interference
components by incorporating the interference taps in a wide range
for more accurate equalization even though the complexity of the
equalizer is high in the early stage of the demodulation and
decoding process.
Here, the equalizer 110 may more accurately estimate interference
using the output of the decoder 130 through the iterative
demodulation and decoding process.
Accordingly, the equalizer 110 may decrease the equalization
complexity while acquiring sufficient equalization performance even
if J, which is used to determine the range of the FTN interference
tap coefficients incorporated therein, is decreased.
Here, the equalizer 110 may decrease J, which is used to determine
the range of the FTN interference tap coefficients incorporated
therein, until J becomes 0.
Further, when J is 0, the equalizer 110 may perform the same
function as the M-ary demodulator.
That is, the equalizer 110 has equalization performance that
improves with the iteration of the demodulation and decoding
process, thereby effectively receiving the FTN interference
signal.
Also, when the result of elimination of the estimated FTN
interference sequence from the FTN signal sequence y satisfies a
preset condition, the decoder 130 may output the decoded LLR
sequence L( ) as an information bit sequence .
Here, when the result of elimination of the estimated FTN
interference sequence from the FTN signal sequence y does not
satisfy the preset condition, the decoder 130 may repeat the
demodulation and decoding process by providing the decoded LLR
sequence L( ) to the interleaver 140.
Here, the demodulation and decoding process may be repeated until
the difference between a.sub.l and {hacek over (a)}.sub.l becomes 0
or until the difference becomes less than a preset value.
FIGS. 5 and 6 are block diagrams that show an FTN-based signal
reception apparatus using a BCJR equalizer and an LDPC decoder
according to an embodiment of the present invention.
Referring to FIG. 5, in an FTN-based signal reception apparatus
using a BCJR equalizer and an LDPC decoder according to an
embodiment of the present invention, the equalizer 110 of the
FTN-based signal reception apparatus illustrated in FIG. 3 may
correspond to the BCJR equalizer 111 for implementing the
Bahl-Cocke-Jelinek-Raviv algorithm, and the decoder 130 may
correspond to the LDPC decoder for performing Low-Density
Parity-Check decoding.
Each of the BCJR equalizer 111 and the LDPC decoder 131 may use an
extrinsic information sequence, acquired by subtracting an input
LLR sequence from an output LLR sequence.
Here, the BCJR equalizer 111 and the LDPC decoder 131 may use the
extrinsic information sequence for stable convergence of an LLR
value depending on the iterative demodulation and decoding
process.
The FTN interference estimation unit 150 is the same as the FTN
interference estimation unit 150 illustrated in FIG. 3, and may use
the FTN filter tap coefficients in Equation (8) in order to set the
interference tap coefficients, incorporated in the BCJR equalizer
111 for the reconstruction of symbols, to `0`.
Also, the FTN interference estimation unit 150 may improve the
equalization performance of the BCJR equalizer 111 by eliminating
the FTN interference components, caused due to the FTN interference
tap coefficients that are not incorporated in the BCJR equalizer
111, using the symbol sequence estimated through the iterative
demodulation and decoding process, and may receive a more accurate
FTN signal by improving the error correction performance of the
LDPC decoder 131.
However, because the FTN-based signal reception apparatus using the
BCJR equalizer 111 and the LDPC decoder 131 requires both a first
interleaver 141 for providing an extrinsic information sequence of
the decoder to the BCJR equalizer 111 and a second interleaver 142
for providing the LDPC-decoded LLR sequence to the FTN interference
estimation unit 150, it is problematic in that the amount of
resources for interleavers is increased when the apparatus is
implemented in hardware.
Referring to FIG. 6, in order to solve the above problem, an
FTN-based signal reception apparatus using the BCJR equalizer 111
and the LDPC decoder 131 in which the structure for calculating an
extrinsic information sequence is changed is shown.
That is, the deinterleaver 120 may deinterleave an extrinsic
information sequence of the equalizer, which is acquired by
eliminating the extrinsic information sequence of the decoder,
calculated by the interleaver, from the LLR sequence created by the
BCJR equalizer 111.
Here, values stored after being calculated in the previous
iteration are used as the extrinsic information sequence of the
decoder, and the initial values in the extrinsic information
sequence of the decoder may be set to `0`.
The interleaver 140 may provide the BCJR equalizer 111 with an
extrinsic information sequence of the decoder, acquired by
eliminating the extrinsic information sequence of the equalizer
from the LLR sequence interleaved after being LDPC-decoded by the
LDPC decoder 131.
Here, because the interleaver 140 may also provide the decoded LLR
sequence, which is LDPC-decoded by the LDPC decoder 131, to the FTN
interference estimation unit 150, only one interleaver 140 is used,
thus reducing the amount of resources for the interleaver 140,
compared to the FTN-based signal reception apparatus using the BCJR
equalizer 111 and the LDPC decoder 131, illustrated in FIG. 5, when
it is implemented in hardware.
Also, the extrinsic information sequence of the decoder may be
calculated using the input and output of the LDPC decoder 131 at
the current iteration. Here, the extrinsic information sequence of
the decoder may be applied as a priori probability in the
equalization process of the BCJR equalizer 111 at the next
iteration.
FIG. 7 is a flowchart that shows a method for receiving a signal
based on FTN according to an embodiment of the present invention.
FIG. 8 is a flowchart that specifically shows an example of the
step of eliminating an FTN interference sequence, illustrated in
FIG. 7.
Referring to FIG. 7, in the method for receiving a signal based on
FTN according to an embodiment of the present invention, an LLR
sequence may be created at step S210.
That is, at step S210, the equalizer 110 may reconstruct distorted
symbols by equalizing an FTN signal sequence y, which is sampled at
the Faster-Than-Nyquist (FTN) signaling rate by being output from a
reception filter, using a given FTN interference pattern, and may
create L({tilde over (v)}), which is a Log Likelihood Ratio (LLR)
sequence of soft information about the received bits, by
demodulating the reconstructed symbols.
Here, L({tilde over (v)}) be the LLR sequence of the received bits
corresponding to v, which is the bit sequence output from the
interleaving process illustrated in FIG. 1.
Also, at step S210, the equalizer 110 may use a variable
interference tap, which changes J, which is used to determine the
range of the FTN interference tap coefficients, depending on the
number of iterations of a demodulation and decoding process.
Here, at step S210, the equalizer 110 may incorporate the
interference tap in a wide range for more accurate equalization in
the early stage of the demodulation and decoding process even if
the complexity of the equalizer is high.
Here, the equalizer 110 may more accurately estimate interference
using the output of the decoder 130 through the iterative
demodulation and decoding process.
Accordingly, at step S210, the equalization complexity may be
decreased while obtaining sufficient equalization performance even
if J, which is used to determine the range of the FTN interference
tap coefficients incorporated in the equalizer 110, is
decreased.
Here, at step S210, J, which is used to determine the range of the
FTN interference tap coefficients incorporated in the equalizer
110, may be decreased until it becomes 0.
Further, at step S210, when J is 0, the equalizer 110 may perform
the same function as an M-ary demodulator.
That is, at step S210, the equalization performance of the
equalizer 110 improves with the iteration of the demodulation and
decoding process, and thus an FTN interference signal may be
effectively received.
Also, at step S210, when the equalizer 110 implements the BCJR
algorithm, the received FTN signal sequence is equalized through
the BCJR algorithm, whereby the LLR sequence L({tilde over (v)})
may be created.
Also, in the method for receiving a signal based on FTN according
to an embodiment of the present invention, deinterleaving may be
performed at step S220.
That is, at step S220, the deinterleaver 120 performs a
deinterleaving process for restoring the order in which elements of
the LLR sequence L({tilde over (v)}), created by the equalizer 110,
are arranged, through the reverse process of interleaving at the
transmitter, and may thereby output the deinterleaved LLR sequence
L({tilde over (c)}).
Also, at step S220, when the equalizer 110 of the FTN-based signal
reception apparatus illustrated in FIG. 3 is a BCJR equalizer 111
for implementing the BCJR algorithm and when the decoder 130 is an
LDPC decoder for performing LDPC decoding, the deinterleaver 120
may deinterleave the extrinsic information sequence of the
equalizer, acquired by eliminating the extrinsic information
sequence of the decoder, calculated by the interleaver, from the
LLR sequence created by the BCJR equalizer 111.
The extrinsic information sequence of the decoder may use values
stored after being calculated in the previous iteration, and the
initial values in the extrinsic information sequence of the decoder
may be set to 0.
Also, in the method for receiving a signal based on FTN according
to an embodiment of the present invention, decoding may be
performed at step S230.
That is, at step S230, the decoder 130 may output a decoded LLR
sequence L( ) by correcting errors in the deinterleaved LLR
sequence L({tilde over (c)}) through a channel decoding
process.
Also, at step S230, when the decoder 130 is the LDPC decoder 131,
the decoder may perform LDPC-decoding on the deinterleaved bit
sequence acquired by deinterleaving the extrinsic information
sequence of the equalizer, and may output the LLR sequence.
Also, in the method for receiving a signal based on FTN according
to an embodiment of the present invention, interleaving may be
performed at step S240.
That is, at step S240, the interleaver 140 performs an interleaving
process on the decoded LLR sequence L( ) in the same order as the
interleaving process at the transmitter, and may thereby output an
interleaved LLR sequence L({hacek over (v)}).
Also, at step S240, the interleaver 140 may provide the BCJR
equalizer 111 with an extrinsic information sequence of the
decoder, acquired by eliminating the extrinsic information sequence
of the equalizer from the LLR sequence, interleaved after being
LDPC-decoded by the LDPC decoder 131.
Here, at step S240, because the interleaver 140 may also provide
the decoded LLR sequence, which is LDPC-decoded by the LDPC decoder
131, to the FTN interference estimation unit 150, only one
interleaver 140 is used, thus reducing the amount of resources for
the interleaver 140 compared to the FTN-based signal reception
apparatus using the BCJR equalizer 111 and the LDPC decoder 131,
illustrated in FIG. 5, when it is implemented in hardware.
Also, the extrinsic information sequence of the decoder may be
calculated using the input and output of the LDPC decoder 131 at
the current iteration. Here, the extrinsic information sequence of
the decoder may be applied as a priori probability in the
equalization process of the BCJR equalizer 111 at the next
iteration.
Also, in the method for receiving a signal based on FTN according
to an embodiment of the present invention, an FTN interference
sequence may be eliminated at step S250.
Referring to FIG. 8, first, a symbol sequence may be estimated at
step S251.
That is, at step S251, the modulation unit 151 of the FTN
interference estimation unit 150 may estimate a symbol sequence
{hacek over (a)} by modulating the interleaved LLR sequence
L({hacek over (v)}) in the same manner as the modulation process at
the transmitter.
Here, at step S251, the modulation unit 151 may use an M-ary
modulation method, and may estimate the length of the symbol
sequence using m=log.sub.2M.
Also, the FTN interference sequence may be estimated at step
S252.
That is, at step S252, the FTN interference filter 152 of the FTN
interference estimation unit 150 calculates the FTN interference
components caused due to the FTN interference filter tap
coefficients that are not incorporated in the equalizer 110, and
thereby acquires the estimated FTN interference sequence shown in
Equation (7).
Here, at step S252, as shown in Equation (7), the FTN interference
filter 152 may estimate the FTN interference sequence by performing
convolution of the symbol sequence {hacek over (a)} estimated by
the modulation unit 151 with {hacek over (g)}, in which the FTN
interference tap coefficients, incorporated in the equalizer in
order to reconstruct symbols, are set to `0`.
Also, the FTN interference sequence may be eliminated at step
S253.
That is, at step S253, the FTN interference filter 152 eliminates
the estimated FTN interference sequence from the received FTN
signal sequence y, thereby creating the FTN signal sequence from
which the estimated FTN interference sequence is eliminated, that
is, y- .
Here, the equation for the k-th element of the FTN signal sequence
from which the estimated interference sequence is eliminated may be
represented as Equation (9).
The expression <1> in Equation (9) is a part that may be
reconstructed using the FTN interference tap coefficients
incorporated in the equalizer 110, and the expression <2> is
a part in which the FTN interference component caused due to the
FTN interference tap coefficients that are not incorporated in the
equalizer 110 is eliminated using the estimated symbol sequence.
The expression <2> may approach 0 a.sub.l becomes equal to
{hacek over (a)}.sub.l. Here, {hacek over (a)}.sub.l is an
estimated symbol, modulated by the modulation unit 151, and if the
result of channel decoding is accurate, {hacek over (a)}.sub.l,
becomes equal to a.sub.l. Accordingly, when the result of channel
decoding becomes accurate through the iterative demodulation and
decoding process, the expression <2> may converge on 0.
In other words, at step S253, through the process of eliminating
the estimated FTN interference sequence , acquired through FTN
interference estimation in the iterative demodulation and decoding
structure, from the FTN signal sequence y sampled at the FTN
signaling rate, the FTN interference estimation unit 150 may
eliminate the interference that cannot be incorporated in the
equalizer 110 due to a complexity problem.
Also, an FTN signal sequence from which the FTN interference
sequence is eliminated may be provided at step S254.
That is, at step S254, because the FTN interference filter 152
provides the equalizer 110 with the FTN signal sequence from which
the FTN interference sequence is eliminated, the equalizer 110 may
have equalization performance that improves with the iteration of
the demodulation and decoding process, and may effectively receive
the FTN interference signal.
Here, at step S254, when the result of elimination of the estimated
FTN interference sequence from the FTN signal sequence y satisfies
a preset condition, the decoder 130 may output the decoded LLR
sequence L( ) as an information bit sequence .
Here, at step S254, when the result of elimination of the estimated
FTN interference sequence from the FTN signal sequence y does not
satisfy the preset condition, the decoder 130 may repeat the
demodulation and decoding process by providing the decoded LLR
sequence L( ) to the interleaver 140.
Here, at step S254, the demodulation and decoding process may be
repeated until the difference between a.sub.l, and {hacek over
(a)}.sub.l, becomes 0 or until the difference becomes less than a
preset value.
The present invention may improve the equalization performance of
an equalizer having relatively low complexity in a digital
communication system using FTN.
Also, the present invention may accurately detect transmitted
signals in an equalization process having relatively low complexity
in a digital communication system using FTN.
Also, the present invention may effectively receive FTN signals by
eliminating interference components that cannot be cancelled in an
equalizer due to a complexity problem.
Also, the present invention may reduce the amount of resources for
an interleaver when implementing a signal receiver using a
Bahl-Cocke-Jelinek-Raviv (BCJR) equalizer and a Low-Density
Parity-Check (LDPC) decoder.
Also, the present invention may improve the equalization
performance of a BCJR equalizer by eliminating FTN interference
components caused due to interference tap coefficients that are not
incorporated in the BCJR equalizer, and may accurately receive FTN
signals by improving the error-correction capability of an LDPC
decoder.
As described above, the apparatus and method for receiving a signal
based on FTN according to the present invention are not limitedly
applied to the configurations and operations of the above-described
embodiments, but all or some of the embodiments may be selectively
combined and configured, so that the embodiments may be modified in
various ways.
* * * * *